Intelligent reflecting surface-assisted group selection beamforming method

Abstract

The present invention relates to the technical field of wireless communications, and in particular to an intelligent reflecting surface-assisted group selection beamforming method. First, a raccoon position is initialized, and an RIS phase shift matrix is selected. Then, users are grouped on the basis of spatial correlation, and a beamforming matrix and an overall system rate are computed. The RIS phase shift matrix is updated on the basis of an obtained result by means of a raccoon optimization algorithm. The proposed algorithm is iterated multiple times to obtain the maximum overall system rate and a corresponding beamforming matrix. The intelligent reflecting surface-assisted group selection beamforming method provided by the present invention achieves joint optimization of the RIS phase shift matrix and the beamforming matrix by introducing the raccoon optimization algorithm, and solves the problems of slow convergence and susceptibility to local optima of other algorithms. By grouping the users when computing the beamforming matrix, the computational complexity of the system is reduced while inter-user interference is reduced, thereby improving the overall system rate.

Description

A Smart Reflector-Assisted Group Selection Beamforming Method Technical Field

[0001] This invention relates to the field of wireless communication technology, and more specifically to a smart reflector-assisted group selection beamforming method. Background Technology

[0002] With the widespread adoption of mobile devices, the extensive use of the internet, and the rapid development of the Internet of Things (IoT), the demand for data transmission rates is constantly increasing. Millimeter waves, due to their shorter wavelength, allow a large number of antenna elements to be packed into a smaller physical area, thus promoting the use of large-scale multiple-input multiple-output (MIMO) systems. Millimeter wave communication offers higher spectral capacity and data transmission rates; however, millimeter wave signals are susceptible to atmospheric absorption, attenuation, and multipath effects during transmission, requiring beamforming technology to overcome these challenges and ensure communication quality and reliability.

[0003] Beamforming is a crucial technology in wireless communication systems, enhancing signal transmission and reception in specific directions to improve system performance and efficiency. In the millimeter-wave band, antenna arrays are typically very large, making direct digital beamforming cumbersome and costly. Hybrid beamforming, combining digital and analog beamforming, can maintain good performance while reducing hardware complexity and cost. However, in certain situations, obstacles or multipath effects in signal propagation can limit traditional beamforming techniques, leading to signal quality degradation or increased communication instability.

[0004] To overcome these challenges, smart reflector technology has been introduced into communication systems. A smart reflector is a novel wireless communication technology that utilizes numerous reflective elements to adjust and optimize the transmission path of electromagnetic waves. This allows for signal control and management without increasing transmission power or using more complex antenna structures, thereby improving signal coverage, enhancing signal strength, and reducing interference.

[0005] By introducing intelligent reflector technology, communication links between users and base stations can be established in environments where line-of-sight propagation paths are obstructed. Traditionally, line-of-sight propagation paths can be blocked by obstacles such as buildings and terrain, resulting in signal attenuation and multipath effects, thus reducing communication quality and coverage. However, the introduction of intelligent reflectors allows for optimization of signal transmission paths in these obstructed environments. By controlling the direction and intensity of signal reflection, obstacles in line-of-sight propagation are effectively overcome, thereby achieving a reliable communication link between users and base stations. The application of this technology can not only improve communication coverage but also enhance signal quality, reduce interference and attenuation in communication, and provide users with a more stable and high-speed wireless communication experience. This paper proposes an intelligent reflector-assisted group selection beamforming technique. The Raccoon Optimization Algorithm is used to jointly optimize the phase shift matrix and beamforming matrix of the intelligent reflector, and group selection technology is used to further reduce the computational complexity of beamforming, thereby maximizing system performance and rate.

[0006] Patent CN115102592B, entitled "A Multi-User MIMO Beamforming Method Based on Federated Learning," provides a beamforming method based on federated learning technology. It uses a base station (BS) and users to jointly train a CNN model. Users train their local models using local channel data and precoder labels. The BS aggregates the local model parameters of all users to obtain a global model, thus achieving beamforming. Patent CN114598368B, entitled "A Full-Duplex Robust Hybrid Beamforming Method Based on Millimeter-Wave Broadband Communication," provides a full-duplex hybrid beamforming method. It utilizes the null-space characteristics of the equivalent channel to eliminate self-interference, then uses the relationship between the lower bound of the channel capacity and WMMSE to solve for the corresponding all-digital transceiver, and finally uses an iterative decomposition method with closed-form solutions to obtain the hybrid transceiver.

[0007] Summary of the Invention

[0008] The purpose of this invention is to propose a smart reflector-assisted group selective beamforming method for optimizing signal transmission and reception in wireless communication systems.

[0009] The intelligent reflector-assisted group selection beamforming method proposed in this invention includes the following steps:

[0010] Step 1: Generate phase-shift codebook F1 and beamforming codebook F based on the resolution τ of the smart reflector and the corresponding vector a(φ) of the antenna array. A ;

[0011] Step 2: Determine the population size S, the dimension of the decision variables D, and the maximum number of iterations T for the raccoon optimization algorithm;

[0012] Step 3: Initialize the raccoon's position m based on the resolution τ of the smart reflective surface;

[0013] Step 4: Generate the phase shift matrix Θ based on the phase shift codebook F1 and the obtained raccoon position m;

[0014] Step 5: Group users according to the spatial correlation of the channel and calculate the beamforming matrix W;

[0015] Step 6: Calculate the total rate R of the system based on the beamforming matrix, and take the population with the largest total rate value as the optimal solution.

[0016] Step 7: Update the iguana's position based on the raccoon's optimal position;

[0017] Step 8: Update the raccoon's location based on the iguana's location and the raccoon's hunting and attack strategy using the optimized algorithm.

[0018] Step 9: Repeat steps 4 to 6 to obtain the optimal individual position of the raccoon during the hunting and attack phase;

[0019] Step 10: Update the raccoon's position according to the predator escape strategy of the raccoon optimization algorithm;

[0020] Step 11: Repeat steps 4 to 6 to obtain the optimal location of the raccoon population;

[0021] Step 12: Repeat steps 7 to 11 until the maximum number of iterations T is reached, to obtain the optimal position m* and the beamforming matrix W corresponding to the optimal position. * This is the desired optimal beamforming matrix.

[0022] Furthermore, in step 1, the generated phase-shift codebook F1 and beamforming codebook F A as follows:

[0023] Furthermore, in step 2, the population size S = 50, the decision variable dimension D is the number of reflective elements N of the intelligent reflective surface, and the maximum number of iterations T = 40.

[0024] Furthermore, in step 3, the initial position m of the raccoon is generated by randomly generating N integers with values ​​ranging from 1 to τ.

[0025] Furthermore, in step 4, the phase shift matrix Θ is generated using the formula Θ=diag(θ1,…,θ n ,…,θ N (3)

[0026] Where, diag(θ1,…,θ) n ,…,θ N) is based on (θ1,…,θ n ,…,θ N ) is a diagonal matrix with the main diagonal as the main diagonal.

[0027] Furthermore, in step 5, N r K represents the number of antennas, K represents the number of users, and the beamforming matrix is... The maximum number of groups is The channel model is specifically as follows: the channel between the k-th user and the base station uses... Indicates that the channel between the k-th user and the smart reflector is used for... This indicates that the channel between the smart reflector and the base station uses... The equivalent channel from the k-th user to the base station can be represented as: h k =h d,k +GΘh r,k .

[0028] Furthermore, in step 6, the formula for calculating the total system speed R is as follows:

[0029] Where, γ k The signal-to-interference-plus-noise ratio (SIR) for the k-th user is expressed as follows:

[0030] Where, p k For transmission power, This represents noise power.

[0031] The raccoon optimization algorithm described above mainly includes the following:

[0032] The initialization phase includes generating the raccoon's initial position. Since codebook selection is an integer programming problem, the raccoon's initial position is a randomly generated integer within the search space.

[0033] During the hunting and attack phase, the optimal population position from the initialization phase is set to the iguana's location. The first half of the raccoon population climbs the trees to hunt the iguana, while the second half waits for the iguana to be startled and fall to the ground. After the iguana falls to the ground, it is placed at a random location in the search space. Based on this random location, the raccoons on the ground move within the search space.

[0034] In the predator escape phase, when a predator attacks a raccoon, the raccoon uses its current location as a reference point to search for a safe location and moves towards it. The determination of the safe location depends on the current iteration number.

[0035] The above user grouping strategy mainly includes the following:

[0036] The channel correlation coefficient is calculated based on the user's channel spatial correlation, forming a correlation coefficient matrix and transforming it into a triangular matrix. The calculation formula is as follows:

[0037] Among them, h i and h j These are the channel vectors for the i-th and j-th users, respectively.

[0038] First, select the two users with the highest correlation from the correlation matrix and group them together, represented as: (v1,v2)=argmax i,j [C] i,j (7)

[0039] Assuming there are m users in the current group, we will assign user v i The correlation threshold is defined as v i The average correlation with other users is calculated using the following formula:

[0040] Then, select users v from the remaining users who are in the same group. i The correlation is greater than the threshold users and form a candidate set If the candidate set P = {t1, t2, ..., t} j Given multiple elements, we select the user with the highest total relevance to the users within the group as the (m+1)th user in this group. Then:

[0041] Repeat the above process until the candidate set P is empty, at which point the grouping ends. Beneficial effects:

[0042] This invention proposes a smart reflector-assisted group selection beamforming method. The basic idea is to decompose the beamforming problem into two sub-problems: RIS phase shift optimization and beamforming matrix optimization, using an alternating optimization approach. The proposed method first initializes the raccoon positions and selects the RIS phase shift matrix. Next, users are grouped based on spatial correlation, and the beamforming matrix and total system rate are calculated. Then, the raccoon optimization algorithm updates the RIS phase shift matrix based on the results. After multiple iterations, the proposed algorithm obtains the maximum total system rate and the corresponding beamforming matrix. This invention's smart reflector-assisted group selection beamforming method, by introducing the raccoon optimization algorithm, achieves joint optimization of the RIS phase shift matrix and the beamforming matrix, solving the problems of slow convergence speed and susceptibility to local optima found in other algorithms. Grouping users reduces interference between users and lowers the system's computational complexity. Test results show that the proposed smart reflector-assisted group selection beamforming method is feasible and universal, with faster convergence speed and higher total system rate. With the help of RIS, it effectively solves the beamforming problem and significantly improves the total system rate. Attached Figure Description

[0043] Figure 1 is a system composition block diagram of the present invention.

[0044] Figure 2 is a flowchart of a group selection beamforming method assisted by an intelligent reflector surface according to the present invention.

[0045] Figure 3 is a simulation diagram of the convergence speed of the algorithm in the embodiment of the present invention.

[0046] Figure 4 is a simulation diagram showing the relationship between signal-to-noise ratio and total system speed in an embodiment of the present invention. Detailed Implementation

[0047] The embodiments of the present invention are described in detail below. These embodiments are implemented based on the technical methods of the present invention, and provide detailed implementation methods and specific operation processes. However, the scope of protection of the present invention is not limited to the following embodiments.

[0048] A smart reflector-assisted group selection beamforming method includes the following steps:

[0049] Step 1: Generate phase-shift codebook F1 and beamforming codebook F based on the resolution τ of the smart reflector and the corresponding vector a(φ) of the antenna array. A ;

[0050] Step 2: Determine the population size S, the dimension of the decision variables D, and the maximum number of iterations T for the raccoon optimization algorithm;

[0051] Step 3: Initialize the raccoon's position m based on the resolution τ of the smart reflective surface;

[0052] Step 4: Generate the phase shift matrix Θ based on the phase shift codebook F1 and the obtained raccoon position m;

[0053] Step 5: Group users according to the spatial correlation of the channel and calculate the beamforming matrix W;

[0054] Step 6: Calculate the total rate R of the system based on the beamforming matrix, and take the population with the largest total rate value as the optimal solution.

[0055] Step 7: Update the iguana's position based on the raccoon's optimal position;

[0056] Step 8: Update the raccoon's location based on the iguana's location and the raccoon's hunting and attack strategy using the optimized algorithm.

[0057] Step 9: Repeat steps 4 to 6 to obtain the optimal individual position of the raccoon during the hunting and attack phase;

[0058] Step 10: Update the raccoon's position according to the predator escape strategy of the raccoon optimization algorithm;

[0059] Step 11: Repeat steps 4 to 6 to obtain the optimal location of the raccoon population;

[0060] Step 12: Repeat steps 7 to 11 until the maximum number of iterations T is reached, and the optimal position m is obtained. * The beamforming matrix W corresponding to the optimal position * This is the desired optimal beamforming matrix.

[0061] Furthermore, in step 1, the resolution τ of the smart reflector is 4, and the generated phase-shift codebook F1 and beamforming codebook F... A as follows:

[0062] in,

[0063] Where λ is the carrier wavelength, d is the antenna spacing, and φ is the angle of arrival.

[0064] Furthermore, in step 2, the population size S = 50, the dimension of the decision variables D = 60, and the maximum number of iterations T = 40.

[0065] Furthermore, in step 3, the initial position m of the raccoon is generated by randomly generating 60 integers from the range of 1 to 4.

[0066] Furthermore, in step 4, the phase shift matrix Θ is generated using the formula Θ=diag(θ1,…,θ n ,…,θ N (13)

[0067] Where, diag(θ1,…,θ) n ,…,θ N ) is based on (θ1,…,θ n ,…,θ N ) is a diagonal matrix with the main diagonal as the main diagonal.

[0068] Furthermore, in step 5, the number of antennas N r =64, number of users K=8, beamforming matrix W=[w1,…,w k ,…,w K The dimension of the [structure] is 64×8, and the maximum number of groups is 4. The channel model is specifically as follows: the channel between the k-th user and the base station uses [the following structure / method]... Indicates that the channel between the k-th user and the smart reflector is used for... This indicates that the channel between the smart reflector and the base station uses... The equivalent channel from the k-th user to the base station can be represented as: h k =h d,k +GΘh r,k .

[0069] Furthermore, in step 6, the formula for calculating the total system speed R is as follows:

[0070] Where, γ k The signal-to-interference-plus-noise ratio (SIR) for the k-th user is expressed as follows:

[0071] Where, p k For transmission power, This represents noise power.

[0072] The raccoon optimization algorithm described above mainly includes the following:

[0073] The initialization phase includes generating the initial positions of the raccoons. Since codebook selection is an integer programming problem, the initial positions of the raccoons are randomly generated integers within the search space. The formula for generating the initial position of the j-th decision vector of the i-th population is as follows: m i,j =round(r·(ub) j -lb j ))+lb j (16)

[0074] Among them ub j and lb j These are the upper and lower bounds of the decision variable, respectively, and ub is set. j =4, lb j=1. r is a random number between 0 and 1. The round(·) function rounds the value. The raccoon's position represents an index in the phase shift codebook; the phase shift matrix Θ can be obtained based on the index corresponding to the raccoon's position. By grouping users and calculating the beamforming matrix, the total system rate in the current state can be obtained.

[0075] During the hunting and attack phase, the position m of the population with the highest traditional total rate in the initialization phase is determined. best The location of the iguana is set, and the first half of the raccoon population climbs the tree to hunt the iguana, while the second half waits for the iguana to be startled and fall to the ground. The formula for updating the location of the raccoons climbing the tree is:

[0076] Where I is a random integer in the set {1,2}.

[0077] After the iguana lands on the ground, it is placed at a random location in the search space. Based on this random location, the raccoon on the ground moves within the search space. The iguana's location is generated using the following formula: G j =round(r·(ub) j -lb j ))+lb j (18)

[0078] The formula for updating the raccoon's position on the ground is as follows:

[0079] Among them, F (·) It is the fitness of the objective function at the corresponding position, i.e., the total system speed.

[0080] In the predator escape phase, when a predator attacks the raccoon, the raccoon uses its current position as a reference point to search for a safe location and moves towards it. The determination of the safe location depends on the current iteration number. The position update formula for this phase is as follows:

[0081] Where t is the current iteration number.

[0082] During both the hunting / attack phase and the escape from predator phase, the raccoon's location is updated according to the following formula:

[0083] Repeat the above process until the maximum number of iterations is reached to obtain the optimal beamforming matrix W. * And the corresponding maximum total system speed.

[0084] The above user grouping strategy mainly includes the following:

[0085] The channel correlation coefficient is calculated based on the user's channel spatial correlation, forming a correlation coefficient matrix and transforming it into a triangular matrix. The calculation formula is as follows:

[0086] Among them, h i and h j These are the channel vectors for the i-th and j-th users, respectively.

[0087] First, select the two users with the highest correlation from the correlation matrix and group them together, represented as: (v1,v2)=argmax i,j [C] i,j (twenty four)

[0088] Assuming there are m users in the current group, we will assign user v i The correlation threshold is defined as v i The average correlation with other users is calculated using the following formula:

[0089] Then, select users v from the remaining users who are in the same group. i The correlation is greater than the threshold users and form a candidate set If the candidate set P = {t1, t2, ..., t} j Given multiple elements, we select the user with the highest total relevance to the users within the group as the (m+1)th user in this group. Then:

[0090] Repeat the above process until the candidate set P is empty, at which point the grouping ends.

[0091] Figure 3 is a simulation diagram of the convergence speed of the algorithm in the embodiment of the present invention. In order to demonstrate the convergence characteristics of the proposed method, the convergence speed of the proposed method is compared with that of the classical particle swarm algorithm. As can be seen from the figure, the convergence speed of the proposed method is faster, and the overall system speed is higher than that of the traditional particle swarm algorithm during convergence. It is also less likely to get trapped in local optima.

[0092] Figure 4 is a simulation diagram showing the relationship between signal-to-noise ratio (SNR) and overall system speed in an embodiment of the present invention. To intuitively demonstrate the superiority of this method, it is compared with a codebook-based greedy algorithm and a grouping algorithm without intelligent reflector assistance. As can be seen from the figure, the overall system speed of this method is superior to the comparison algorithms under all five SNR conditions. At an SNR of 20 dB, the overall system speed of this method is 22% higher than the grouping algorithm without intelligent reflector assistance and 56% higher than the codebook-based greedy algorithm. The intelligent reflector-assisted grouping selection beamforming method proposed in this invention effectively solves the beamforming problem with the assistance of RIS (Resonance Analysis System) and significantly improves the overall system speed.

Claims

1. An intelligent reflecting surface-assisted grouped selection beamforming method, characterized in that, The application method includes the following steps: Step 1, generating phase shift codebook F1 and beamforming codebook F2 according to the resolution τ of the smart reflecting surface and the corresponding vector a(φ) of the antenna array A ; Step 2: Determine the population size S, the dimension of the decision variables D, and the maximum number of iterations T for the raccoon optimization algorithm; Step 3: Initialize the raccoon's position m based on the resolution τ of the smart reflective surface; Step 4: Generate the phase shift matrix Θ based on the phase shift codebook F1 and the obtained raccoon position m; Step 5: Group users according to the spatial correlation of the channel and calculate the beamforming matrix W; Step 6: Calculate the total rate R of the system based on the beamforming matrix, and take the population with the largest total rate value as the optimal solution. Step 7: Update the iguana's position based on the raccoon's optimal position; Step 8: Update the raccoon's location based on the iguana's location and the raccoon's hunting and attack strategy using the optimized algorithm. Step 9: Repeat steps 4 to 6 to obtain the optimal individual position of the raccoon during the hunting and attack phase; Step 10: Update the raccoon's position according to the predator escape strategy of the raccoon optimization algorithm; Step 11: Repeat steps 4 to 6 to obtain the optimal location of the raccoon population; Step 12, repeat steps 7 to 11 until the maximum iteration number T is reached to obtain the optimal position m * the beamforming matrix W corresponding to the optimal position * is the optimal beamforming matrix 2. The smart reflector-assisted grouped selection beamforming method of claim 1, wherein, In step 2, the population size S of the raccoon optimization algorithm is 50, the dimension of the decision variable D is the number of reflective elements N of the smart reflective surface, and the maximum number of iterations T is 40.

3. The intelligent reflector-assisted grouped selective beamforming method according to claim 1, characterized in that, In step 3, the initial position m of the raccoon is generated by randomly generating N integers from 1 to τ.

4. The intelligent reflector-assisted grouped selective beamforming method according to claim 1, characterized in that, In step 4, the generation of the phase shift matrix Θ, the specific formula is Θ = diag(θ1, …, θN) where θ1, …, θNare the phase shifts of the N antennas. n ,…,θ N In step 4, the generation of the phase shift matrix Θ, the specific formula is Θ = diag(θ1, …, θN) where where diag(θ1,..., θ n ,...,θ N ) is a diagonal matrix with (θ1,..., θ n ,...,θ N ) as the main diagonal.

5. The intelligent reflector-assisted grouped selective beamforming method according to claim 1, characterized in that, In step 5, N r K represents the number of antennas, K represents the number of users, and the beamforming matrix is... Maximum number of groups is The channel model is specifically as follows: the channel between the k-th user and the base station uses... Indicates that the channel between the k-th user and the smart reflector is used for... This indicates that the channel between the smart reflector and the base station uses... The equivalent channel from the k-th user to the base station can be represented as: h k =h d,k +GΘh r,k .

6. The intelligent reflector-assisted grouped selective beamforming method according to claim 1, characterized in that, In step 6, the formula for calculating the total system speed R is: Where, γ k Let be the signal-to-interference-plus-noise ratio (SIR) for the k-th user.

Images